There are 3 major forms of degenerative dementia; Alzheimer disease (AD), Dementia with Lewy bodies (DLB), and Frontotemporal Dementias (FTD). Convergent clinical, pathologic, and genetic data indicate that these degenerative dementias have different mechanisms of pathogenesis. It is likely that different treatments will be needed to alter the course of these disorders. Appropriate application of treatments will require precise and early diagnosis. Precise, early diagnosis would be very useful now to improve clinical trial performance. Improved precision of initial diagnosis would also be valuable for present clinical practice. Certain differentiation of these entities is possible with postmortem evaluations only. Clinical criteria exist to guide ante-mortem diagnosis but these criteria are imprecise. Differentiating DLB and FTD from AD is particularly difficult, especially in individuals with mild-early dementia. Imaging methods may improve the precision of initial diagnosis. We found that quantifying nigrostriatal dopamine terminals robustly differentiates AD and DLB. FTD may be differentiated from AD and DLB by distinctive patterns of regional cortical hypometabolism. We have developed a method, [11C]dihydrotetrabenazine positron emission tomography (DTBZ-PET), that quantifies nigrostriatal terminal density and assesses regional cortical perfusion as a faithful analogue of regional cortical metabolism. Our preliminary data suggests that this method will permit early, accurate differentiation of AD, DLB, and FTD. Our goal is to develop an imaging method to permit early and accurate differentiatal diagnosis of early mild dementia. We propose a prospective evaluation of the ability of DTBZ-PET to differentiate the 3 major degenerative dementias in subjects with early-mild dementia. Our hypothesis is that the results of DTBZ-PET imaging in subjects with mild, early dementia will predict their evolution into clinically distinguished AD, FTD, and DLB. Our first specific aim is to determine nigrostriatal terminal density and patterns of regional cortical perfusion in a cohort of well characterized subjects with early-mild dementia. Our second specific aim is respectively to follow and characterize this cohort with standardized clinical and psychometric evaluations. Our third specific aim is to correlate the clinically and psychometrically established diagnoses with diagnoses predicted by DTBZ-PET. If successful, these studies would strongly support the use of DTBZ-PET imaging or similar methods in the evaluation of early dementia.